6.1 Encoding: Transforming Perceptions into Memories

Bubbles P., a professional gambler with no formal education, who spent most of his time shooting craps at local clubs or playing high-stakes poker, had no difficulty rattling off 20 numbers, in either forward or backward order, after just a single glance (Ceci, DeSimone, & Johnson, 1992). Most people can listen to a list of numbers and then repeat them from memory–as long as the list is no more than about seven items long (try it for yourself using FIGURE 6.1).

Figure 6.1: Digit Memory Test How many digits can you remember? Start on the first row and cover the rows below it with a piece of paper. Study the numbers in the row for 1 second and then cover that row back up again. After a couple of seconds, try to repeat the numbers. Then uncover the row to see if you were correct. If so, continue down to the next row, using the same instructions, until you can’t recall all the numbers in a row. The number of digits in the last row you can remember correctly is your digit span. Bubbles P. could remember 20 random numbers, or about five rows deep. How did you do?

How did Bubbles accomplish his astounding feats of memory? For at least 2,000 years, people have thought of memory as a recording device that makes exact copies of information that comes in through our senses, and then stores those copies for later use. This idea is simple and intuitive. It is also completely incorrect. Memories are made by combining information we already have in our brains with new information that comes in through our senses. In this way memory is like cooking; starting from a recipe but improvising along the way, we add old information to new information, mix, shake, bake, and out pops a memory. Memories are constructed, not recorded, and encoding is the process by which we transform what we perceive, think, or feel into an enduring memory. Let’s look at three types of encoding processes–semantic encoding, visual imagery encoding, and organizational encoding–and then consider the possible survival value of encoding for our ancestors.

Semantic Encoding

Have you ever wondered why you can remember 20 experiences (your favorite camping trip, your 16th birthday party, your first day at college, etc.) but not 20 digits? One reason is that we often think about the meaning behind our experiences, so we semantically encode them without even trying (Craik & Tulving, 1975).

ISTOCKPHOTO/THINKSTOCK

Memories are a combination of old and new information, so the nature of any particular memory depends as much on the old information already in our memories as it does on the new information coming in through our senses. In other words, how we remember something depends on how we think about it at the time. For example, as a professional gambler, Bubbles found numbers unusually meaningful, so when he saw a string of digits, he tended to think about their meanings. He might have thought about how they related to his latest bet at the racetrack or to his winnings after a long night at the poker table. Whereas you might try to memorize the string 22061823 by saying it over and over, Bubbles would think about betting $220 at 6-to-1 odds on horse number 8 to place 2nd in the 3rd race. Indeed, when Bubbles was tested with materials other than numbers–faces, words, objects, or locations–his memory performance was no better than average.

In one study, researchers presented participants with a series of words and asked them to make one of three types of judgments (Craik & Tulving, 1975): semantic judgments required the participants to think about the meaning of the words (Is hat a type of clothing?); rhyme judgments required the participants to think about the sound of the words (Does hat rhyme with cat?); and visual judgments required the participants to think about the appearance of the words (Is HAT written uppercase or lowercase?). The type of judgment task influenced how participants thought about each word–what old information they combined with the new–and had a powerful impact on their memories. Those participants who made semantic judgments (i.e., had thought about the meaning of the words) had much better memory for the words than did participants who thought about how the word looked or sounded. The results of these and many other studies have shown that long-term retention is greatly enhanced by semantic encoding, the process of relating new information in a meaningful way to knowledge that is already stored in memory (Brown & Craik, 2000).

So where does this semantic encoding take place? What’s going on in the brain when this type of information processing occurs? Studies reveal that semantic encoding is uniquely associated with increased activity in the lower left part of the frontal lobe and the inner part of the left temporal lobe (FIGURE 6.2a; Demb et al., 1995; Kapur et al., 1994; Wagner et al., 1998). In fact, the amount of activity in each of these two regions during encoding is directly related to whether people later remember an item. The more activity there is in these areas, the more likely the person will remember the information.

Visual Imagery Encoding

In Athens in 477 BCE, the Greek poet Simonides had just left a banquet when the ceiling collapsed and killed all the people inside. Simonides was able to name every one of the dead simply by visualizing each chair around the banquet table and recalling the person who had been sitting there. Simonides wasn’t the first, but he was among the most proficient, to use visual imagery encoding, the process of storing new information by converting it into mental pictures.

Semantic encoding involves relating new information in a meaningful way to facts you already know; visual imagery encoding involves storing new information by converting it into mental pictures. How might you use both kinds of encoding to help store a new fact, such as the date of a friend’s birthday that falls on, say, November 24th?

RADIUS IMAGES/ALAMY

If you wanted to use Simonides’ method to create an enduring memory, you could simply convert the information that you wanted to remember into a visual image and then store it in a familiar location. For instance, if you were going to the grocery store and wanted to remember to buy Coke, popcorn, and cheese dip, you could use the rooms in your house as locations and imagine your living room flooded in Coke, your bedroom pillows stuffed with popcorn, and your bathtub as a greasy pond of cheese dip. When you arrived at the store, you could then take a mental walk around your house and “look” into each room to remember the items you needed to purchase.

Numerous experiments have shown that visual imagery encoding can substantially improve memory. In one experiment, participants who studied lists of words by creating visual images of them later recalled twice as many items as participants who just mentally repeated the words (Schnorr & Atkinson, 1969). Why does visual imagery encoding work so well? First, visual imagery encoding does some of the same things that semantic encoding does: When you create a visual image, you relate incoming information to knowledge already in memory. For example, a visual image of a parked car might help you create a link to your memory of your first kiss.

Second, when you use visual imagery to encode words and other verbal information, you end up with two different mental placeholders for the items–a visual one and a verbal one–which gives you more ways to remember them than just a verbal placeholder alone (Paivio, 1971, 1986). Visual imagery encoding activates visual processing regions in the occipital lobe (see FIGURE 6.2b), which suggests that people actually enlist the visual system when forming memories based on mental images (Kosslyn et al., 1993).

Figure 6.2: Brain Activity during Different Types of Judgments fMRI studies reveal that different parts of the brain are active during different types of judgments: (a) During semantic judgments, the lower left frontal lobe is active; (b) during visual judgments, the occipital lobe is active; and (c) during organizational judgments, the upper left frontal lobe is active.

Organizational Encoding

Have you ever ordered dinner with a group of friends and watched in amazement as your server took the order without writing anything down? To find out how this is done, one researcher spent 3 months working in a restaurant where servers routinely wrote down orders but then left the check at the customer’s table before proceeding to the kitchen and telling the cooks what to make (Stevens, 1988). The researcher wired the servers with microphones and asked them to think aloud, that is, to say what they were thinking as they walked around all day doing their jobs. The researcher found that as soon as the server left a customer’s table, he or she immediately began grouping or categorizing the orders into hot drinks, cold drinks, hot foods, and cold foods. The servers grouped the items into a sequence that matched the layout of the kitchen, first placing drink orders, then hot food orders, and finally cold food orders. The servers remembered their orders by relying on organizational encoding, the process of categorizing information according to the relationships among a series of items.

Ever wonder how a server remembers who ordered the pizza and who ordered the fries without writing anything down? Some have figured out how to use organizational encoding.

JEFF GREENBERG/ALAMY

For example, suppose you had to memorize the words peach, cow, chair, apple, table, cherry, lion, couch, horse, desk. The task seems difficult, but if you organize the items into three categories–fruit (peach, apple, cherry), animals (cow, lion, horse), and furniture (chair, table, couch, desk)–the task becomes much easier. Studies have shown that instructing people to sort items into categories like this is an effective way to enhance their subsequent recall of those items (Mandler, 1967). Even more complex organizational schemes have been used, such as the hierarchy in FIGURE 6.3 (Bower et al., 1969). People can improve their recall of individual items by organizing them into multiple-level categories, all the way from a general category such as animals, through intermediate categories such as birds and songbirds, down to specific examples such as wren and sparrow.

Figure 6.3: Organizing Words into a Hierarchy Organizing words into conceptual groups and relating them to one another–such as in this example of a hierarchy–makes it easier to reconstruct the items from memory later (Bower et al., 1969). Keeping track of the 17 items in this example can be facilitated by remembering the hierarchical groupings they fall under.

Just as semantic and visual imagery encoding activates distinct regions of the brain, so, too, does organizational encoding. As you can see in FIGURE 6.2c, organizational encoding activates the upper surface of the left frontal lobe (Fletcher, Shallice, & Dolan, 1998; Savage et al., 2001). Different types of encoding strategies appear to rely on different areas of brain activation.

Encoding of Survival-Related Information

Encoding new information is critical to many aspects of everyday life–prospects for attaining your degree would be pretty slim without this ability–and the survival of our ancestors likely depended on encoding and later remembering such things as the sources of food and water or where a predator appeared (Nairne & Pandeirada, 2008; Sherry & Schacter, 1987).

Recent experiments have addressed these ideas by examining encoding of survival-related information. The experiments were motivated by an evolutionary perspective based on Darwin’s principle of natural selection: The features of an organism that help it survive and reproduce are more likely than other features to be passed on to subsequent generations (see the Psychology: Evolution of a Science chapter). Therefore, memory mechanisms that help us to survive and reproduce should be preserved by natural selection, and our memory systems should be built in a way that allows us to remember especially well encoded information that is relevant to our survival.

To test this idea, the researchers gave participants three different encoding tasks (Nairne, Thompson, & Pandeirada, 2007). In the survival-encoding condition, participants were asked to imagine that they were stranded in the grasslands of a foreign land without any survival materials and that over the next few months they would need supplies of food and water and also need to protect themselves from predators. The researchers then showed participants randomly chosen words (e.g., stone, meadow, chair) and asked them to rate on a 1–5 scale how relevant each item would be to survival in the hypothetical situation. In the moving-encoding condition, a second group of participants was asked to imagine that they were planning to move to a new home in a foreign land, and to rate on a 1–5 scale how useful each item might be in helping them to set up a new home. Finally, in the pleasantness-encoding condition, a third group was shown the same words and asked to rate on a 1–5 scale the pleasantness of each word.

The findings, displayed in FIGURE 6.4, show that participants recalled more words after the survival-encoding task than after either the moving or pleasantness tasks. In later studies, the researchers found that survival encoding resulted in higher levels of recall than several other non-survival-encoding tasks involving semantic encoding, imagery encoding, or organizational encoding (Nairne, Pandeirada, & Thompson, 2008). Exactly what about survival encoding produces such high levels of memory? Survival encoding draws on elements of semantic, visual imagery, and organizational encoding, which may give it an advantage over any one of the other three (Burns, Hwang, & Burns, 2011). Also, survival encoding encourages participants to engage in extensive planning, which in turn benefits memory and may account for much of the benefit of survival encoding. For example, when participants imagine scenarios in which they are stranded in grasslands without food, and encode a series of words with respect to their survival relevance, survival scenarios that do involve planning produce superior subsequent memory compared with survival scenarios that do not involve planning. Critically, superior recall is also observed for scenarios that involve planning but not survival, such as planning a dinner party (Klein, Robertson, & Delton, 2011). Of course, planning for the future is itself critical for our long-term survival, so these findings are still broadly consistent with an evolutionary perspective in which memory is built to support planning and related forms of thinking about the future that enhance our chances of survival (Klein et al., 2011; Schacter, 2012; Suddendorf & Corballis, 2007).

Figure 6.4: Survival Encoding Enhances Later Recall What does a pouncing cougar that may threaten our survival have to do with recall? People recall more words after survival encoding (Nairne et al., 2007).

DON JOHNSTON/ALL CANADA PHOTOS/GETTY IMAGES